REPORT ON CHEMISTRY. 
417 
easily determined. But while chemists differ as much in regard 
to the true specific gravities as in regard to the atomic weights, 
it is obvious that to correct the latter by the above equation 
would be to escape one error by combining two others of per¬ 
haps equal amount. The true numerical expressions therefore 
for each series, the specific gravities and the atomic weights, 
must be sought for by separate and distinct experimental re¬ 
searches ; and the relation between them can he employed only 
as a check upon the results thus obtained. It was in reference 
to this point that the Chemical Committee of the British Asso¬ 
ciation, at the Meeting in York, deemed it of high importance 
to draw the attention of chemical philosophers to the necessity 
of a more accurate experimental inquiry into the true specific 
gravities of the simple gases. 
Can we obtain the true atoms , or only multiples of them ?— 
The above remarks have had reference to the minute differences 
between the weights received in this, and in other countries, 
and which differences, in most cases, do not amount to the 
atomic weight of hydrogen. Though these differences are the 
greatest in importance, they are not the only ones to which the 
limited nature of our present knowledge has given rise. It is 
still a question, whether the atomic weights we obtain for each 
simple substance, supposing them correct, represent the true 
weights of the atoms, or only multiples or submultiples of them; 
-—whether, for instance, the weight of hydrogen be *0625 or 
*125, that of mercury 25 or 1 2'5. It is in many cases easy 
enough to come to a probable conclusion on this point, from 
the consideration of .the different compounds which the simple 
substances form with each other; but in some instances the evi¬ 
dence in favour of the greater or the less number is so equally 
balanced as to render the decision a matter of great difficulty, 
and one on which the most eminent chemists disagree. 
An indirect auxiliary to these determinations has been sought 
for in the investigation of the specific heats of bodies. Ever 
since the experiments of Dulong and Petit showed that in the 
case of simple substances there was reason to suppose that the 
specific heats multiplied by the atomic weights give a constant 
quantity, the confirmation of their results by an accurate de¬ 
termination of the specific heats of all bodies, both simple and 
compound, has become of great importance in atomic chemistry. 
It is not probable that we shall ever determine these heats with 
sufficient accuracy to enable us to employ them in discriminating 
among the series of minute differences to which we have above 
adverted ; butan establishment of the law of Dulong and Petit by 
tolerable approximations to the true specific heats, would enable 
2 D 
